EP1032241B1 - Method and system for switching using an arbitrator - Google Patents
Method and system for switching using an arbitrator Download PDFInfo
- Publication number
- EP1032241B1 EP1032241B1 EP00301399A EP00301399A EP1032241B1 EP 1032241 B1 EP1032241 B1 EP 1032241B1 EP 00301399 A EP00301399 A EP 00301399A EP 00301399 A EP00301399 A EP 00301399A EP 1032241 B1 EP1032241 B1 EP 1032241B1
- Authority
- EP
- European Patent Office
- Prior art keywords
- traffic
- arbitrator
- segment
- destinations
- queue
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Expired - Lifetime
Links
Images
Classifications
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L49/00—Packet switching elements
- H04L49/30—Peripheral units, e.g. input or output ports
- H04L49/3081—ATM peripheral units, e.g. policing, insertion or extraction
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04Q—SELECTING
- H04Q11/00—Selecting arrangements for multiplex systems
- H04Q11/04—Selecting arrangements for multiplex systems for time-division multiplexing
- H04Q11/0428—Integrated services digital network, i.e. systems for transmission of different types of digitised signals, e.g. speech, data, telecentral, television signals
- H04Q11/0478—Provisions for broadband connections
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L12/00—Data switching networks
- H04L12/54—Store-and-forward switching systems
- H04L12/56—Packet switching systems
- H04L12/5601—Transfer mode dependent, e.g. ATM
- H04L2012/5678—Traffic aspects, e.g. arbitration, load balancing, smoothing, buffer management
- H04L2012/5679—Arbitration or scheduling
Definitions
- the present invention is related to multi level dequeuing of traffic in switching networks when multiple sources and destinations are involved, with each source having traffic destined to different destinations over mutiple prioritized queues. Any scheme that is fair in resource allocation can be used for selection of different prioritized queues.
- the multi level dequeuing is especially useful for high bandwidth network applications, which makes it feasible to implement scalable switches with current technology.
- the switch design involves collecting and queuing of incoming traffic, for example packets or asynchronous transfer mode (ATM) cells, from many sources, for example physical media. Each source could have traffic destined for different destinations over different priorities. Incoming traffic is queued based on either per destination or per destination and per priority queue. As the number of sources and destinations increase, and/or as each source and destination traffic bandwidth increases it becomes increasingly difficult to build a switching module that can switch by queuing and dequeuing traffic at very high bandwidths. The technology limitations would make it impossible to build switches beyond certain bandwidth with single level of queuing and dequeing. US 5,745,489 is an example of such a prior art switching system.
- the purpose of the present invention is to build network switches that can scale to any bandwidth by using multi node queuing and multi level dequeuing.
- Each switch comprises one or more queuing/dequeuing segments which each run at only a fraction of the total bandwidth and an arbitrator arranged to collect information from all queues of each lower level dequeuing segment and by processing the information from different segments based on a certain algorithm, for example a weighted round robin algorithm, it will give back dequeuing information to each segment.
- the arbitrator needs to receive only enough information from all its lower level segments for a current dequeuing interval. For each lower level segment the arbitrator can receive information from both the queuing logic and the destination of the traffic.
- the approach can be extended to any number of dequeuing levels by giving enough time to transfer the data from the buffer queues to the destinations.
- a switch for switching traffic from N sources to M destinations comprises K segments, where K is greater than or equal to 2 and is an integer, each segment being arranged to receive traffic from R of the N sources, where 1 ⁇ R ⁇ N and is an integer, all K segments in total being arranged to receive traffic from the N sources, and each segment is arranged to collect and queue traffic from the respective R sources, and an arbitrator arranged to receive information from the destinations regarding if they can receive data or not, and from the K segments about the traffic they have for different destinations.
- Each segment may include input ports, each having a line rate L1, which receive traffic from corresponding sources, the arbitrator may include output ports each having a line rate L2 and a combination of an input port and an output port together may form a queue/dequeue node having a bandwidth of N/K * L1 + M * L2.
- Each segment may have at least one queue associated with a destination and in which each is arranged to store traffic received at its input ports for its corresponding destinations.
- Each segment may have a controller which places traffic received at an input port into a queue corresponding with the traffic's destinations.
- the traffic in a queue may either be a unicast connection type or a multicast connection type.
- the arbitrator may comprise a arbitrator controller arranged to select which destination is to receive traffic from a corresponding source and allows the traffic to flow through an output port to a destination.
- the arbitrator may select a destination, then an associated queue having traffic for that destination and dequeue the traffic in the associated queue to the destination through an output port.
- the arbitrator may have dequeuing intervals in which traffic is sent to destinations and wherein each segment sends the arbitrator buffer occupancy information for all destinations and queues for each dequeuing interval of the arbitrator.
- the arbitrator may comprise arbitrator portions, each arbitrator portion associated with predetermined segments and the arbitrator portions and segments form a hierarchy. The arbitrator may select a destination according to either a weighted round robin or a strict round robin routine.
- the queues are priority queues.
- a method of switching traffic from N sources to M destinations includes receiving traffic from the N sources at input ports of K segments, where K is greater than or equal to 2 and is an integer, each segment receiving traffic from R of the N sources where 1 ⁇ R ⁇ N and is an integer, and sending traffic from any segment to a destination through an output port of an arbitrator which connects to each segment.
- the method may include receiving traffic at an input port having a line rate of L1 and sending traffic from an output port of the arbitrator at a line rate of L2, wherein the combination of an input port and an output port together form a queue/dequeue node having a bandwidth of N/K * L1 + M * L2.
- the method may include placing traffic received at an input port into a queue in the segment corresponding with the traffic destination. It may also include, after the receiving traffic, selecting a destination using the arbitrator. Also it may including, after selecting a destination, selecting a queue within a segment associated with the selected destination. Furthermore, it may include, after selecting a queue, selecting a unicast or multicast connection within the queue. The method may also include, after selecting a connection, selecting a segment with traffic for the connection.
- an arbitrator arranged to receive information from M destinations regarding if they can receive data or not arising from N sources, and from K segments about the traffic they have for different destinations arising from the N sources, each segment having input ports having a line rate L1, where K, M and N are each an integer greater than or equal to 2, comprises output ports each having a line rate L2 and a combination of an input port and an output port together form a queue/dequeue node having a bandwidth of N/K * L1+ M * L2, and an arbitrator controller arranged to select which destination is to receive traffic from a corresponding source and to allow the traffic to flow through an output port to the destination.
- the controller may define the queuing intervals in which traffic is sent to destinations and the controller receives buffer occupancy information from each segment for all destinations.
- the controller may select a destination according to either a weighted round robin or a strict round robin routine.
- a switch 10 for switching traffic from N sources 12 to M destinations 14, where M and N are each an integer greater than or equal to 2.
- the switch 10 comprises K segments 16, where K is greater than or equal to 2 and is an integer.
- Each segment 16 receives traffic from R of the N sources 12, where 1 ⁇ R ⁇ N and is an integer, and all K segments 16 in total receive traffic from the N sources 12.
- Each segment 16 collects and queues traffic from the respective R sources 12.
- the switch 10 comprises an arbitrator 18 which receives information from the destinations 14 regarding if they can receive data or not, and from the K segments 16 about the traffic they have for different destinations 14.
- each segment 16 includes input ports 20, each having a line rate L1, which receive traffic from corresponding sources 12.
- the arbitrator 18 includes output ports 22 each having a line rate L2, and a combination of an input port 20 and an output port 22 together form a queue/dequeue node having a bandwidth of N/K * L1 + M * L2.
- Each segment 16 preferably has queues 24 associated with destinations 14 which store traffic received at the input ports 20 for the destinations 14.
- each segment 16 has a controller 26 which places traffic received at an input port 20 into a queue 24 corresponding with the traffic's destinations 14.
- Traffic in a queue 24 is preferably either a unicast connection type or a multicast connection type.
- the arbitrator 18 selects a destination and then an associated queue 24 having traffic for that destination 14, and dequeues the traffic in the associated queue 24 to the destination through an output port 22.
- the arbitrator 18 preferably has dequeuing intervals in which traffic is sent to destinations 14 and wherein each segment 16 sends the arbitrator 18 buffer occupancy information for all destinations 14 and queues 24 for each dequeuing interval of the arbitrator 18.
- the arbitrator 18 comprises arbitrator portions 28, each arbitrator portion 28 associated with a predetermined segment 16 and the arbitrator portions 28 and segments 16 form a hierarchy.
- the arbitrator 18 preferably selects a destination 14 according to either a weighted round robin or a strict round robin routine.
- the queues 24 are priority queues 24.
- the present invention also pertains to a method for switching traffic from N sources 12 to M destinations 14, where M and N are each an integer greater than or equal to 2.
- the method comprises receiving traffic from the N sources 12 at input ports 20 of K segments 16, where K is greater than or equal to 2 and is an integer.
- K is greater than or equal to 2 and is an integer.
- Each segment 16 receives traffic from R of the N sources 12 where 1 ⁇ R ⁇ N and is an integer. Then traffic is sent from any segment 16 to a destination through an output port 22 of an arbitrator 18 which connects to each segment 16.
- traffic is received at an input port 20 having a line rate of L1 and traffic is sent from an output port 22 of the arbitrator 18 at a line rate of L2.
- a combination of an input port 20 and an output port 22 together form a queue/dequeue node having a bandwidth of N/K * L1 + M * L2.
- traffic received at an input port 20 is placed in a queue 24 in the segment 16 corresponding with the traffic destination.
- a destination 14 and a queue 24 associated with the destination 14 is selected by the arbitrator 18. After a queue 24 has been selected, the arbitrator 18 selects either a unicast or a multicast connection within the queue 24 and selects a segment 16 with traffic for the connection.
- the present invention also pertains to an arbitrator 18 which receives information from M destinations 14 regarding if they can receive data or not arising from N sources 12, and from K segments 16 about the traffic they have for different destinations 14 arising from the N sources 12.
- Each segment 16 has input ports 20 having a line rate L1, where K, M and N are each an integer greater than or equal to 2.
- the arbitrator 18 comprises output ports 22 each having a line rate L2, and a combination of an input port 20 and an output port 22 together form a queue/dequeue node having a bandwidth of N/K * L1+ M * L2.
- the arbitrator 18 comprises an arbitrator controller 43 that selects which destination 14 is to receive traffic from a corresponding source and allows the traffic to flow through an output port 22 to the destination 14.
- the arbitrator controller 43 defines the queuing intervals in which traffic is sent to destinations 14 and the arbitrator controller 43 receives buffer occupancy information from each segment 16 for all destinations.
- the arbitrator controller 43 preferably selects a destination according to either a weighted round robin or a strict round robin routine.
- the present invention pertains to a segment 16 which receives traffic from R of N sources 12, where 1 is less than or equal to R which is less than N and is an integer, and N is an integer greater than or equal to 2.
- the segment 16 comprises input ports 20, each having a line rate L1, which receive traffic from the respective R sources 12.
- the segment 16 comprises queues 24 associated with destinations 14 and are arranged to store traffic received at the input ports 20 for the destinations 14.
- the segment includes a controller 26 which places traffic received at an input port 20 into a queue 24 corresponding with the traffic's destinations.
- Traffic in a queue 24 can preferably be either a unicast connection type or a multicast connection type.
- the queues 24 are priority queues.
- the arbitrator 18 can be extended to any number of levels.
- Traffic queued in different prioritized queues 24 from each input port 20 can go to any one (unicast connections) or more (multicast connections) of the M output ports 22.
- each of the two segments 16 collects and queues the traffic from N/2 sources 12. This is in contrast to single point queuing and dequeuing where it is necessary to queue and dequeue traffic from N sources 12.
- the queuing approach used in this instance is per destination and per priority queue for a destination.
- the arbitrator 18 receives information from the destinations 14 regarding if they can receive data or not, and from the two segments 16 about the buffer occupancy for different destinations 14 and priority queues 24. Each segment 16 can send the arbitrator 18 only enough information for the current dequeuing interval about its buffer occupancy for all destinations 14 and queues 24. The arbitrator 18 selects the destination 16 first, then a queue 24 associated with that destination 16, and then one of the two segments 16, using, for example, a weighted round-robin routine at each level. Once the arbitrator 18 makes the final decision it will send the required dequeue command to the appropriate segment 16.
- Any fair algorithm can be used to implement the dequeue algorithm within the arbitrator 18.
- the algorithm has to be fair in allocating the bandwidth not only across destinations 14 and priority queues 24 associated with each destination, but across the different segments 16.
- a novel algorithm for an arbitrator 18 design is explained below with reference to Figure 3 .
- a destination is selected using a strict round-robin routine.
- a destination can be selected using a weighted round-robin routine. If a selected destination does not have traffic to be dequeued then that dequeue interval could be wasted. Accordingly, after selecting a destination, a priority queue 24 associated with that destination is selected using a weighted round-robin routine. If the selected queue does not have traffic to be dequeued in all of the lower level segments 16, then the queue 24 next in priority will be selected until a queue 24 with traffic queued is found. After selecting a priority queue, a unicast or multicast connection type is selected within that priority queue 24, based on a weighted round-robin routine.
- connection type does not have traffic queued in all of the lower level segments 16 then another connection type will be selected. Multiple weights can be used in selecting a connection type, based on the queuing algorithm used for queuing incoming traffic, and the current congestion state of all lower level segments 16. After selecting a connection type, one of the lower level segments 16 is selected using a strict round-robin routine until a segment 16 with traffic queued for the selected destination, priority queue, and connection type is found.
- the memory bandwidth required for ATM cell queuing from N ports and ATM cell dequeuing to M ports could be prohibitively high to implement it as a single point queuing and dequeuing.
- the current memory technology may not make it feasible to implement such single point queuing and dequeuing logic.
- the queuing could be segmented and using an arbitrator 18 approach dequeuing can be extended to any number of levels as required, making it feasible to implement a scalable switching node that can scale to any aggregate bandwidth.
Abstract
Description
- The present invention is related to multi level dequeuing of traffic in switching networks when multiple sources and destinations are involved, with each source having traffic destined to different destinations over mutiple prioritized queues. Any scheme that is fair in resource allocation can be used for selection of different prioritized queues. The multi level dequeuing is especially useful for high bandwidth network applications, which makes it feasible to implement scalable switches with current technology.
- In high speed switching networks, the switch design involves collecting and queuing of incoming traffic, for example packets or asynchronous transfer mode (ATM) cells, from many sources, for example physical media. Each source could have traffic destined for different destinations over different priorities. Incoming traffic is queued based on either per destination or per destination and per priority queue. As the number of sources and destinations increase, and/or as each source and destination traffic bandwidth increases it becomes increasingly difficult to build a switching module that can switch by queuing and dequeuing traffic at very high bandwidths. The technology limitations would make it impossible to build switches beyond certain bandwidth with single level of queuing and dequeing.
US 5,745,489 is an example of such a prior art switching system. - The purpose of the present invention is to build network switches that can scale to any bandwidth by using multi node queuing and multi level dequeuing. Each switch comprises one or more queuing/dequeuing segments which each run at only a fraction of the total bandwidth and an arbitrator arranged to collect information from all queues of each lower level dequeuing segment and by processing the information from different segments based on a certain algorithm, for example a weighted round robin algorithm, it will give back dequeuing information to each segment. The arbitrator needs to receive only enough information from all its lower level segments for a current dequeuing interval. For each lower level segment the arbitrator can receive information from both the queuing logic and the destination of the traffic. The approach can be extended to any number of dequeuing levels by giving enough time to transfer the data from the buffer queues to the destinations.
- According to a first aspect of the invention a switch for switching traffic from N sources to M destinations, where M and N are each an integer greater than or equal to 2, comprises K segments, where K is greater than or equal to 2 and is an integer, each segment being arranged to receive traffic from R of the N sources, where 1 ≤ R < N and is an integer, all K segments in total being arranged to receive traffic from the N sources, and each segment is arranged to collect and queue traffic from the respective R sources, and an arbitrator arranged to receive information from the destinations regarding if they can receive data or not, and from the K segments about the traffic they have for different destinations.
- Each segment may include input ports, each having a line rate L1, which receive traffic from corresponding sources, the arbitrator may include output ports each having a line rate L2 and a combination of an input port and an output port together may form a queue/dequeue node having a bandwidth of N/K * L1 + M * L2. Each segment may have at least one queue associated with a destination and in which each is arranged to store traffic received at its input ports for its corresponding destinations. Each segment may have a controller which places traffic received at an input port into a queue corresponding with the traffic's destinations.
- The traffic in a queue may either be a unicast connection type or a multicast connection type.
- The arbitrator may comprise a arbitrator controller arranged to select which destination is to receive traffic from a corresponding source and allows the traffic to flow through an output port to a destination. The arbitrator may select a destination, then an associated queue having traffic for that destination and dequeue the traffic in the associated queue to the destination through an output port. The arbitrator may have dequeuing intervals in which traffic is sent to destinations and wherein each segment sends the arbitrator buffer occupancy information for all destinations and queues for each dequeuing interval of the arbitrator. The arbitrator may comprise arbitrator portions, each arbitrator portion associated with predetermined segments and the arbitrator portions and segments form a hierarchy. The arbitrator may select a destination according to either a weighted round robin or a strict round robin routine.
- Preferably, the queues are priority queues.
- According to a second aspect of the invention a method of switching traffic from N sources to M destinations, where M and N are each an integer greater than or equal to 2, includes receiving traffic from the N sources at input ports of K segments, where K is greater than or equal to 2 and is an integer, each segment receiving traffic from R of the N sources where 1 ≤ R < N and is an integer, and sending traffic from any segment to a destination through an output port of an arbitrator which connects to each segment.
- The method may include receiving traffic at an input port having a line rate of L1 and sending traffic from an output port of the arbitrator at a line rate of L2, wherein the combination of an input port and an output port together form a queue/dequeue node having a bandwidth of N/K * L1 + M * L2. The method may include placing traffic received at an input port into a queue in the segment corresponding with the traffic destination. It may also include, after the receiving traffic, selecting a destination using the arbitrator. Also it may including, after selecting a destination, selecting a queue within a segment associated with the selected destination. Furthermore, it may include, after selecting a queue, selecting a unicast or multicast connection within the queue. The method may also include, after selecting a connection, selecting a segment with traffic for the connection.
- According to a further aspect of the invention an arbitrator arranged to receive information from M destinations regarding if they can receive data or not arising from N sources, and from K segments about the traffic they have for different destinations arising from the N sources, each segment having input ports having a line rate L1, where K, M and N are each an integer greater than or equal to 2, comprises output ports each having a line rate L2 and a combination of an input port and an output port together form a queue/dequeue node having a bandwidth of N/K * L1+ M * L2, and an arbitrator controller arranged to select which destination is to receive traffic from a corresponding source and to allow the traffic to flow through an output port to the destination.
- Preferably, the controller may define the queuing intervals in which traffic is sent to destinations and the controller receives buffer occupancy information from each segment for all destinations. The controller may select a destination according to either a weighted round robin or a strict round robin routine.
- According to a further aspect of the invention there is provided a segment suitable for use with a switch.
- In the accompanying drawings, the preferred embodiment of the invention and preferred methods of practicing the invention are illustrated, by way of example only, in which:
-
Figure 1 is a schematic representation of a switch according to the present invention. -
Figure 2 is a schematic representation of a switch according to the present invention. -
Figure 3 is a flow chart of a method according to the present invention. - Referring now to the drawings, wherein like reference numerals refer to similar or identical parts throughout the several views, and more specifically to
Figure 1 thereof, there is shown aswitch 10 for switching traffic fromN sources 12 toM destinations 14, where M and N are each an integer greater than or equal to 2. Theswitch 10 comprisesK segments 16, where K is greater than or equal to 2 and is an integer. Eachsegment 16 receives traffic from R of theN sources 12, where 1 ≤ R < N and is an integer, and allK segments 16 in total receive traffic from theN sources 12. Eachsegment 16 collects and queues traffic from therespective R sources 12. Theswitch 10 comprises anarbitrator 18 which receives information from thedestinations 14 regarding if they can receive data or not, and from theK segments 16 about the traffic they have fordifferent destinations 14. - Preferably, each
segment 16 includesinput ports 20, each having a line rate L1, which receive traffic fromcorresponding sources 12. Preferably, thearbitrator 18 includesoutput ports 22 each having a line rate L2, and a combination of aninput port 20 and anoutput port 22 together form a queue/dequeue node having a bandwidth of N/K * L1 + M * L2. Eachsegment 16 preferably hasqueues 24 associated withdestinations 14 which store traffic received at theinput ports 20 for thedestinations 14. - Preferably, each
segment 16 has acontroller 26 which places traffic received at aninput port 20 into aqueue 24 corresponding with the traffic'sdestinations 14. Traffic in aqueue 24 is preferably either a unicast connection type or a multicast connection type. Preferably, thearbitrator 18 selects a destination and then an associatedqueue 24 having traffic for thatdestination 14, and dequeues the traffic in the associatedqueue 24 to the destination through anoutput port 22. - The
arbitrator 18 preferably has dequeuing intervals in which traffic is sent todestinations 14 and wherein eachsegment 16 sends thearbitrator 18 buffer occupancy information for alldestinations 14 andqueues 24 for each dequeuing interval of thearbitrator 18. Preferably, thearbitrator 18 comprisesarbitrator portions 28, eacharbitrator portion 28 associated with apredetermined segment 16 and thearbitrator portions 28 andsegments 16 form a hierarchy. - The
arbitrator 18 preferably selects adestination 14 according to either a weighted round robin or a strict round robin routine. Preferably, thequeues 24 arepriority queues 24. - The present invention also pertains to a method for switching traffic from
N sources 12 toM destinations 14, where M and N are each an integer greater than or equal to 2. The method comprises receiving traffic from theN sources 12 atinput ports 20 ofK segments 16, where K is greater than or equal to 2 and is an integer. Eachsegment 16 receives traffic from R of theN sources 12 where 1 ≤ R < N and is an integer. Then traffic is sent from anysegment 16 to a destination through anoutput port 22 of anarbitrator 18 which connects to eachsegment 16. - Preferably, traffic is received at an
input port 20 having a line rate of L1 and traffic is sent from anoutput port 22 of thearbitrator 18 at a line rate of L2. A combination of aninput port 20 and anoutput port 22 together form a queue/dequeue node having a bandwidth of N/K * L1 + M * L2. Preferably traffic received at aninput port 20 is placed in aqueue 24 in thesegment 16 corresponding with the traffic destination. - A
destination 14 and aqueue 24 associated with thedestination 14 is selected by thearbitrator 18. After aqueue 24 has been selected, thearbitrator 18 selects either a unicast or a multicast connection within thequeue 24 and selects asegment 16 with traffic for the connection. - The present invention also pertains to an
arbitrator 18 which receives information fromM destinations 14 regarding if they can receive data or not arising fromN sources 12, and fromK segments 16 about the traffic they have fordifferent destinations 14 arising from the N sources 12. Eachsegment 16 hasinput ports 20 having a line rate L1, where K, M and N are each an integer greater than or equal to 2. Thearbitrator 18 comprisesoutput ports 22 each having a line rate L2, and a combination of aninput port 20 and anoutput port 22 together form a queue/dequeue node having a bandwidth of N/K * L1+ M * L2. Thearbitrator 18 comprises anarbitrator controller 43 that selects whichdestination 14 is to receive traffic from a corresponding source and allows the traffic to flow through anoutput port 22 to thedestination 14. - Preferably, the
arbitrator controller 43 defines the queuing intervals in which traffic is sent todestinations 14 and thearbitrator controller 43 receives buffer occupancy information from eachsegment 16 for all destinations. Thearbitrator controller 43 preferably selects a destination according to either a weighted round robin or a strict round robin routine. - Furthermore, the present invention pertains to a
segment 16 which receives traffic from R ofN sources 12, where 1 is less than or equal to R which is less than N and is an integer, and N is an integer greater than or equal to 2. Thesegment 16 comprisesinput ports 20, each having a line rate L1, which receive traffic from the respective R sources 12. Thesegment 16 comprisesqueues 24 associated withdestinations 14 and are arranged to store traffic received at theinput ports 20 for thedestinations 14. - Preferably, the segment includes a
controller 26 which places traffic received at aninput port 20 into aqueue 24 corresponding with the traffic's destinations. Traffic in aqueue 24 can preferably be either a unicast connection type or a multicast connection type. Preferably, thequeues 24 are priority queues. - In the operation of the invention, and referring to
Figure 2 , the following describes the approach with asingle level arbitrator 18 with twosegments 16. It should be noted that thearbitrator 18 can be extended to any number of levels. Suppose there is traffic coming fromN sources 12 into a network switch node, and after it is appropriately queued and dequeued it goes toM destinations 14. Traffic queued in different prioritizedqueues 24 from eachinput port 20 can go to any one (unicast connections) or more (multicast connections) of theM output ports 22. With one level ofarbitrator 18 each of the twosegments 16 collects and queues the traffic from N/2sources 12. This is in contrast to single point queuing and dequeuing where it is necessary to queue and dequeue traffic from N sources 12. The queuing approach used in this instance is per destination and per priority queue for a destination. Thearbitrator 18, based on a fixed algorithm described in detail below, dequeues for each dequeue interval. - The
arbitrator 18 receives information from thedestinations 14 regarding if they can receive data or not, and from the twosegments 16 about the buffer occupancy fordifferent destinations 14 andpriority queues 24. Eachsegment 16 can send thearbitrator 18 only enough information for the current dequeuing interval about its buffer occupancy for alldestinations 14 andqueues 24. Thearbitrator 18 selects thedestination 16 first, then aqueue 24 associated with thatdestination 16, and then one of the twosegments 16, using, for example, a weighted round-robin routine at each level. Once thearbitrator 18 makes the final decision it will send the required dequeue command to theappropriate segment 16. For example, if theinput ports 20 have a line rate of L1, and theoutput ports 22 have a line rate of L2, then the bandwidth capability required at each queue/dequeue node with single point dequeuing and multi level dequeuing as proposed above would be: -
-
-
- This clearly shows that with segmented queuing/dequeuing each queue/dequeue node need only support a fraction of the total bandwidth, which makes it feasible to build high speed scalable switching networks. Also, with a
multi level arbitrator 18, thearbitrator 18 design can be made simpler. - Any fair algorithm can be used to implement the dequeue algorithm within the
arbitrator 18. The algorithm has to be fair in allocating the bandwidth not only acrossdestinations 14 andpriority queues 24 associated with each destination, but across thedifferent segments 16. A novel algorithm for anarbitrator 18 design is explained below with reference toFigure 3 . - If all
destinations 14 have the same bandwidth capacity then a destination is selected using a strict round-robin routine. Alternatively, a destination can be selected using a weighted round-robin routine. If a selected destination does not have traffic to be dequeued then that dequeue interval could be wasted. Accordingly, after selecting a destination, apriority queue 24 associated with that destination is selected using a weighted round-robin routine. If the selected queue does not have traffic to be dequeued in all of thelower level segments 16, then thequeue 24 next in priority will be selected until aqueue 24 with traffic queued is found. After selecting a priority queue, a unicast or multicast connection type is selected within thatpriority queue 24, based on a weighted round-robin routine. If the selected connection type does not have traffic queued in all of thelower level segments 16 then another connection type will be selected. Multiple weights can be used in selecting a connection type, based on the queuing algorithm used for queuing incoming traffic, and the current congestion state of alllower level segments 16. After selecting a connection type, one of thelower level segments 16 is selected using a strict round-robin routine until asegment 16 with traffic queued for the selected destination, priority queue, and connection type is found. - For an ATM switching node with N high
speed input ports 20, and M highspeed output ports 22, the memory bandwidth required for ATM cell queuing from N ports and ATM cell dequeuing to M ports could be prohibitively high to implement it as a single point queuing and dequeuing. The current memory technology may not make it feasible to implement such single point queuing and dequeuing logic. In such a situation, the queuing could be segmented and using anarbitrator 18 approach dequeuing can be extended to any number of levels as required, making it feasible to implement a scalable switching node that can scale to any aggregate bandwidth.
Claims (17)
- A switch (10) for high speed switching networks for switching traffic from N sources (12) to M destinations (14), where M and N are each an integer greater than or equal to 2 comprising:K segments (16), where K is greater than or equal to 2 and is an integer, each segment being arranged to receive traffic from R of the N sources (12), where 1 ≤ R < N and is an integer, and all K segments (16) in total being arranged to receive traffic from the N sources (12), each segment being arranged to collect and queue traffic from the respective R sources (12); andcharacterized byan arbitrator (18) connected to each of the segments (16) and being arranged to receive information from the destinations (14) regarding if they can receive data or not, and from the K segments (16) about the traffic they have for different destinations (14), and for sending traffic from any segment (16) to a destination (14) through an output port (22) of the arbitrator (18).
- A switch (10) as described in Claim 1 characterized by the fact that each segment includes input ports (20), each having a line rate L1, adapted to receive traffic from corresponding sources (12), and the arbitrator (18) includes output ports (22) each having a line rate L2 and a combination of an input port and output port together form a queue/dequeue node having a bandwidth of N/K * L1 + M * L2.
- A switch (10) as described in Claim 2 characterized by the fact that each segment has queues (24) associated with destinations (14) which are arranged to store traffic received at the input ports (20) for the destinations (14).
- A switch (10) as described in Claim 3 characterized by the fact that each segment has a controller (26) adapted to place traffic received at an input port into a queue corresponding with the traffic's destinations (14).
- A switch (10) as described in Claim 4 characterized by the fact that traffic in a queue can be either a unicast connection type or a multicast connection type.
- A switch (10) as described in Claim 5 characterized by the fact that the arbitrator (18) is adapted to select a destination and then an associated queue having traffic for that destination, and to drequeue the traffic in the associated queue with the destination through an output port.
- A switch (10) as described in Claim 6 characterized by the fact that the arbitrator (18) has dequeueing intervals in which traffic is sent to destinations (14) and wherein each segment is adapted to send the arbitrator (18) buffer occupancy information for all destinations (14) and queues (24) for each dequeueing interval of the arbitrator (18).
- A switch (10) as described in Claim 7 characterized by the fact that the arbitrator (18) comprises arbitrator portions (28), each arbitrator portion (28) being associated with predetermined segments (16) and the arbitrator portions (28) and the segments (16) forming a hierarchy.
- A switch (10) as described in Claim 8 characterized by the fact that the arbitrator (18) is adapted to select a destination according to either weighted round robin or strict round robin.
- A switch (10) as described in Claim 9 characterized by the fact that the queues (24) are priority queues (24).
- A method for switching traffic from N sources (12) to M destinations (14) for high speed switching networks where M and N are each an integer greater than or equal to 2, comprising the steps of:receiving traffic from the N sources (12) at input ports (20) of K segments (16), where K is greater than or equal to 2 and is an integer, each segment (16) receiving traffic from R of the N sources (12) where 1. R < N and is an integer; each segment collecting and queueing traffic from the respective R sources and characterized bysending traffic from any segment (16) to a destination (14) through an output port (22) of an arbitrator (18) which connects to each segment (16), the arbitrator (18) receiving information from the destinations regarding if they can receive data or not and from the K segments about the traffic they have for different destinations.
- A methods as described in Claim 11 characterized by the fact that the receiving step includes the step of receiving traffic at an input port having a line rate of L1 and the sending step includes the step of sending traffic from an output port of the arbitrator (18) at a line rate of L2, and a combination of an input port and output port together form a queue/dequeue node having a bandwidth of N/K * L1 + M * L2.
- A method as described in Claim 12 characterized by the fact that the receiving step includes the step of placing traffic received at an input port into a queue in the segment corresponding with the traffic destination.
- A method as described in Claim 13 including after the receiving step there is the step of selecting a destination by the arbitrator (18).
- A method as described in Claim 14 including after the destination selecting step there is the step of selecting a queue within a segment associated with the selected destination.
- A method as described in Claim 15 including after the queue selecting step there is the step of selecting a unicast or multicast connection within the queue.
- A method as described in Claim 16 including after the connection selecting step there is the step of selecting a segment with traffic for the connection.
Applications Claiming Priority (2)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
US255197 | 1999-02-22 | ||
US09/255,197 US6647011B1 (en) | 1999-02-22 | 1999-02-22 | Method and system for switching using an arbitrator |
Publications (3)
Publication Number | Publication Date |
---|---|
EP1032241A2 EP1032241A2 (en) | 2000-08-30 |
EP1032241A3 EP1032241A3 (en) | 2003-12-17 |
EP1032241B1 true EP1032241B1 (en) | 2011-10-05 |
Family
ID=22967266
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
EP00301399A Expired - Lifetime EP1032241B1 (en) | 1999-02-22 | 2000-02-22 | Method and system for switching using an arbitrator |
Country Status (5)
Country | Link |
---|---|
US (1) | US6647011B1 (en) |
EP (1) | EP1032241B1 (en) |
JP (1) | JP2000349788A (en) |
AT (1) | ATE527823T1 (en) |
ES (1) | ES2373033T3 (en) |
Families Citing this family (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US6862265B1 (en) * | 2000-04-13 | 2005-03-01 | Advanced Micro Devices, Inc. | Weighted fair queuing approximation in a network switch using weighted round robin and token bucket filter |
SI1575977T1 (en) * | 2002-12-23 | 2010-01-29 | Dynavax Tech Corp | Immunostimulatory sequence oligonucleotides and methods of using the same |
US7352699B2 (en) * | 2003-04-25 | 2008-04-01 | Alcatel Usa Sourcing, L.P. | Switch fabric access scheduler |
US8098660B2 (en) * | 2005-08-01 | 2012-01-17 | Panasonic Corporation | Transmitting apparatus and transmitting method |
US9626309B1 (en) | 2014-07-02 | 2017-04-18 | Microsemi Storage Solutions (U.S.), Inc. | Method and controller for requesting queue arbitration and coalescing memory access commands |
Family Cites Families (15)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US5303078A (en) * | 1990-12-18 | 1994-04-12 | Bell Communications Research, Inc. | Apparatus and method for large scale ATM switching |
US5361255A (en) * | 1991-04-29 | 1994-11-01 | Dsc Communications Corporation | Method and apparatus for a high speed asynchronous transfer mode switch |
JPH0744544B2 (en) * | 1992-01-17 | 1995-05-15 | 富士通株式会社 | Interconnection network with self-routing function |
GB2288096B (en) * | 1994-03-23 | 1999-04-28 | Roke Manor Research | Apparatus and method of processing bandwidth requirements in an ATM switch |
US5537400A (en) * | 1994-04-15 | 1996-07-16 | Dsc Communications Corporation | Buffered crosspoint matrix for an asynchronous transfer mode switch and method of operation |
JP3291122B2 (en) * | 1994-06-02 | 2002-06-10 | 富士通株式会社 | Self-routing switch, ATM switch and switching system |
US5475679A (en) * | 1994-12-08 | 1995-12-12 | Northern Telecom Limited | Large capacity ATM switch |
US5668798A (en) * | 1995-04-05 | 1997-09-16 | International Business Machines Corporation | Multiplexed TC sublayer for ATM switch |
AU6501496A (en) * | 1995-07-19 | 1997-02-18 | Ascom Nexion Inc. | Point-to-multipoint transmission using subqueues |
JP2827998B2 (en) * | 1995-12-13 | 1998-11-25 | 日本電気株式会社 | ATM exchange method |
JPH09247176A (en) * | 1996-03-11 | 1997-09-19 | Hitachi Ltd | Asynchronous transfer mode exchange system |
US6101183A (en) * | 1996-06-21 | 2000-08-08 | Lucent Technologies Inc. | Method and apparatus for crossconnecting transmission members in the outside distribution plant of a telecommunications network |
US5768257A (en) * | 1996-07-11 | 1998-06-16 | Xylan Corporation | Input buffering/output control for a digital traffic switch |
US5841556A (en) * | 1996-12-26 | 1998-11-24 | Electronics And Telecommunications Research Institute | Optical switching system with combined space and wavelength division multiplex based on the repeated contention resolution technique |
US6163542A (en) * | 1997-09-05 | 2000-12-19 | Carr; David Walter | Virtual path shaping |
-
1999
- 1999-02-22 US US09/255,197 patent/US6647011B1/en not_active Expired - Lifetime
-
2000
- 2000-02-22 JP JP2000105698A patent/JP2000349788A/en active Pending
- 2000-02-22 AT AT00301399T patent/ATE527823T1/en not_active IP Right Cessation
- 2000-02-22 ES ES00301399T patent/ES2373033T3/en not_active Expired - Lifetime
- 2000-02-22 EP EP00301399A patent/EP1032241B1/en not_active Expired - Lifetime
Also Published As
Publication number | Publication date |
---|---|
JP2000349788A (en) | 2000-12-15 |
EP1032241A3 (en) | 2003-12-17 |
EP1032241A2 (en) | 2000-08-30 |
ATE527823T1 (en) | 2011-10-15 |
ES2373033T3 (en) | 2012-01-30 |
US6647011B1 (en) | 2003-11-11 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
EP0785699B1 (en) | Multicast routing in multistage networks | |
EP0886939B1 (en) | Efficient output-request packet switch and method | |
US6188690B1 (en) | Method and apparatus for high speed, scalable communication system | |
JP3866425B2 (en) | Packet switch | |
JP3347926B2 (en) | Packet communication system and method with improved memory allocation | |
EP0785698B1 (en) | Buffering of multicast cells in switching networks | |
EP0763915B1 (en) | Packet transfer device and method adaptive to a large number of input ports | |
EP1573950B1 (en) | Apparatus and method to switch packets using a switch fabric with memory | |
EP0717532A1 (en) | Dynamic fair queuing to support best effort traffic in an ATM network | |
US6603772B1 (en) | Multicast routing with multicast virtual output queues and shortest queue first allocation | |
US6396843B1 (en) | Method and apparatus for guaranteeing data transfer rates and delays in data packet networks using logarithmic calendar queues | |
US20030072326A1 (en) | Method and apparatus for parallel, weighted arbitration scheduling for a switch fabric | |
EP0982970B1 (en) | ATM switch | |
AU736780B2 (en) | Method for providing delays independent of switch size in a crossbar switch with speedup | |
EP2134037A1 (en) | Method and apparatus for scheduling data packet flows | |
US6970466B2 (en) | Packet switching apparatus | |
JPH1168770A (en) | Scheduling system for atm switch | |
US6208652B1 (en) | Increment scheduler | |
EP1032241B1 (en) | Method and system for switching using an arbitrator | |
EP1041772A2 (en) | Packet switch realizing transmission with no packet delay | |
US7269158B2 (en) | Method of operating a crossbar switch | |
US6363069B1 (en) | Complete packet discarding | |
Mhamdi et al. | Practical scheduling algorithms for high-performance packet switches | |
ES2235366T3 (en) | PROCEDURE AND SYSTEM OF CIRCUITS TO TRANSMIT MESSAGE UNITS IN MESSAGE FLOWS OF DIFFERENT PRIORITY. | |
KR100458206B1 (en) | Apparatus and method of weighted round-robin cell scheduling for ATM |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
PUAI | Public reference made under article 153(3) epc to a published international application that has entered the european phase |
Free format text: ORIGINAL CODE: 0009012 |
|
AK | Designated contracting states |
Kind code of ref document: A2 Designated state(s): AT BE CH CY DE DK ES FI FR GB GR IE IT LI LU MC NL PT SE |
|
AX | Request for extension of the european patent |
Free format text: AL;LT;LV;MK;RO;SI |
|
PUAL | Search report despatched |
Free format text: ORIGINAL CODE: 0009013 |
|
AK | Designated contracting states |
Kind code of ref document: A3 Designated state(s): AT BE CH CY DE DK ES FI FR GB GR IE IT LI LU MC NL PT SE |
|
AX | Request for extension of the european patent |
Extension state: AL LT LV MK RO SI |
|
RIC1 | Information provided on ipc code assigned before grant |
Ipc: 7H 04Q 11/04 A Ipc: 7H 04L 12/56 B |
|
RAP1 | Party data changed (applicant data changed or rights of an application transferred) |
Owner name: MARCONI INTELLECTUAL PROPERTY (RINGFENCE) INC. |
|
17P | Request for examination filed |
Effective date: 20040609 |
|
AKX | Designation fees paid |
Designated state(s): AT BE CH CY DE DK ES FI FR GB GR IE IT LI LU MC NL PT SE |
|
17Q | First examination report despatched |
Effective date: 20041214 |
|
RAP1 | Party data changed (applicant data changed or rights of an application transferred) |
Owner name: ERICSSON AB |
|
RAP1 | Party data changed (applicant data changed or rights of an application transferred) |
Owner name: ERICSSON AB |
|
RAP1 | Party data changed (applicant data changed or rights of an application transferred) |
Owner name: ERICSSON AB |
|
GRAJ | Information related to disapproval of communication of intention to grant by the applicant or resumption of examination proceedings by the epo deleted |
Free format text: ORIGINAL CODE: EPIDOSDIGR1 |
|
GRAP | Despatch of communication of intention to grant a patent |
Free format text: ORIGINAL CODE: EPIDOSNIGR1 |
|
GRAS | Grant fee paid |
Free format text: ORIGINAL CODE: EPIDOSNIGR3 |
|
GRAA | (expected) grant |
Free format text: ORIGINAL CODE: 0009210 |
|
AK | Designated contracting states |
Kind code of ref document: B1 Designated state(s): AT BE CH CY DE DK ES FI FR GB GR IE IT LI LU MC NL PT SE |
|
REG | Reference to a national code |
Ref country code: GB Ref legal event code: FG4D |
|
REG | Reference to a national code |
Ref country code: CH Ref legal event code: EP |
|
REG | Reference to a national code |
Ref country code: IE Ref legal event code: FG4D |
|
REG | Reference to a national code |
Ref country code: DE Ref legal event code: R096 Ref document number: 60046498 Country of ref document: DE Effective date: 20111208 |
|
REG | Reference to a national code |
Ref country code: NL Ref legal event code: VDEP Effective date: 20111005 |
|
REG | Reference to a national code |
Ref country code: ES Ref legal event code: FG2A Ref document number: 2373033 Country of ref document: ES Kind code of ref document: T3 Effective date: 20120130 |
|
REG | Reference to a national code |
Ref country code: AT Ref legal event code: MK05 Ref document number: 527823 Country of ref document: AT Kind code of ref document: T Effective date: 20111005 |
|
PG25 | Lapsed in a contracting state [announced via postgrant information from national office to epo] |
Ref country code: BE Free format text: LAPSE BECAUSE OF FAILURE TO SUBMIT A TRANSLATION OF THE DESCRIPTION OR TO PAY THE FEE WITHIN THE PRESCRIBED TIME-LIMIT Effective date: 20111005 |
|
PG25 | Lapsed in a contracting state [announced via postgrant information from national office to epo] |
Ref country code: NL Free format text: LAPSE BECAUSE OF FAILURE TO SUBMIT A TRANSLATION OF THE DESCRIPTION OR TO PAY THE FEE WITHIN THE PRESCRIBED TIME-LIMIT Effective date: 20111005 Ref country code: GR Free format text: LAPSE BECAUSE OF FAILURE TO SUBMIT A TRANSLATION OF THE DESCRIPTION OR TO PAY THE FEE WITHIN THE PRESCRIBED TIME-LIMIT Effective date: 20120106 Ref country code: SE Free format text: LAPSE BECAUSE OF FAILURE TO SUBMIT A TRANSLATION OF THE DESCRIPTION OR TO PAY THE FEE WITHIN THE PRESCRIBED TIME-LIMIT Effective date: 20111005 Ref country code: PT Free format text: LAPSE BECAUSE OF FAILURE TO SUBMIT A TRANSLATION OF THE DESCRIPTION OR TO PAY THE FEE WITHIN THE PRESCRIBED TIME-LIMIT Effective date: 20120206 |
|
PG25 | Lapsed in a contracting state [announced via postgrant information from national office to epo] |
Ref country code: CY Free format text: LAPSE BECAUSE OF FAILURE TO SUBMIT A TRANSLATION OF THE DESCRIPTION OR TO PAY THE FEE WITHIN THE PRESCRIBED TIME-LIMIT Effective date: 20111005 |
|
PG25 | Lapsed in a contracting state [announced via postgrant information from national office to epo] |
Ref country code: DK Free format text: LAPSE BECAUSE OF FAILURE TO SUBMIT A TRANSLATION OF THE DESCRIPTION OR TO PAY THE FEE WITHIN THE PRESCRIBED TIME-LIMIT Effective date: 20111005 |
|
PLBE | No opposition filed within time limit |
Free format text: ORIGINAL CODE: 0009261 |
|
STAA | Information on the status of an ep patent application or granted ep patent |
Free format text: STATUS: NO OPPOSITION FILED WITHIN TIME LIMIT |
|
PG25 | Lapsed in a contracting state [announced via postgrant information from national office to epo] |
Ref country code: IT Free format text: LAPSE BECAUSE OF FAILURE TO SUBMIT A TRANSLATION OF THE DESCRIPTION OR TO PAY THE FEE WITHIN THE PRESCRIBED TIME-LIMIT Effective date: 20111005 |
|
26N | No opposition filed |
Effective date: 20120706 |
|
PG25 | Lapsed in a contracting state [announced via postgrant information from national office to epo] |
Ref country code: MC Free format text: LAPSE BECAUSE OF NON-PAYMENT OF DUE FEES Effective date: 20120229 |
|
REG | Reference to a national code |
Ref country code: CH Ref legal event code: PL |
|
PG25 | Lapsed in a contracting state [announced via postgrant information from national office to epo] |
Ref country code: LI Free format text: LAPSE BECAUSE OF NON-PAYMENT OF DUE FEES Effective date: 20120229 Ref country code: CH Free format text: LAPSE BECAUSE OF NON-PAYMENT OF DUE FEES Effective date: 20120229 |
|
REG | Reference to a national code |
Ref country code: DE Ref legal event code: R097 Ref document number: 60046498 Country of ref document: DE Effective date: 20120706 |
|
REG | Reference to a national code |
Ref country code: IE Ref legal event code: MM4A |
|
PG25 | Lapsed in a contracting state [announced via postgrant information from national office to epo] |
Ref country code: IE Free format text: LAPSE BECAUSE OF NON-PAYMENT OF DUE FEES Effective date: 20120222 Ref country code: AT Free format text: LAPSE BECAUSE OF FAILURE TO SUBMIT A TRANSLATION OF THE DESCRIPTION OR TO PAY THE FEE WITHIN THE PRESCRIBED TIME-LIMIT Effective date: 20111005 |
|
PG25 | Lapsed in a contracting state [announced via postgrant information from national office to epo] |
Ref country code: FI Free format text: LAPSE BECAUSE OF FAILURE TO SUBMIT A TRANSLATION OF THE DESCRIPTION OR TO PAY THE FEE WITHIN THE PRESCRIBED TIME-LIMIT Effective date: 20111005 |
|
PG25 | Lapsed in a contracting state [announced via postgrant information from national office to epo] |
Ref country code: LU Free format text: LAPSE BECAUSE OF NON-PAYMENT OF DUE FEES Effective date: 20120222 |
|
REG | Reference to a national code |
Ref country code: FR Ref legal event code: PLFP Year of fee payment: 17 |
|
PGFP | Annual fee paid to national office [announced via postgrant information from national office to epo] |
Ref country code: DE Payment date: 20160226 Year of fee payment: 17 Ref country code: ES Payment date: 20160226 Year of fee payment: 17 |
|
PGFP | Annual fee paid to national office [announced via postgrant information from national office to epo] |
Ref country code: FR Payment date: 20160217 Year of fee payment: 17 Ref country code: GB Payment date: 20160226 Year of fee payment: 17 |
|
REG | Reference to a national code |
Ref country code: DE Ref legal event code: R119 Ref document number: 60046498 Country of ref document: DE |
|
GBPC | Gb: european patent ceased through non-payment of renewal fee |
Effective date: 20170222 |
|
REG | Reference to a national code |
Ref country code: FR Ref legal event code: ST Effective date: 20171031 |
|
PG25 | Lapsed in a contracting state [announced via postgrant information from national office to epo] |
Ref country code: DE Free format text: LAPSE BECAUSE OF NON-PAYMENT OF DUE FEES Effective date: 20170901 Ref country code: FR Free format text: LAPSE BECAUSE OF NON-PAYMENT OF DUE FEES Effective date: 20170228 |
|
PG25 | Lapsed in a contracting state [announced via postgrant information from national office to epo] |
Ref country code: GB Free format text: LAPSE BECAUSE OF NON-PAYMENT OF DUE FEES Effective date: 20170222 |
|
REG | Reference to a national code |
Ref country code: ES Ref legal event code: FD2A Effective date: 20180507 |
|
PG25 | Lapsed in a contracting state [announced via postgrant information from national office to epo] |
Ref country code: ES Free format text: LAPSE BECAUSE OF NON-PAYMENT OF DUE FEES Effective date: 20170223 |